How Web development technology was shaped
A long time ago, during the 1980s, in an institute called CERN (the European institute for nuclear research, in Switzerland), a British contractor called Tim Berners-Lee, together with some colleagues, came up with an idea to establish a connection via computer and land telephone lines between many researchers spread around the world and the institute, in order to do improve the exchange of academic information.
Some scientific articles from CERN’s researchers were already available on a local network written using some form of markup. It allowed for articles with some library-like indexing as well as basic formatting like headers, basic text marks, titles and not a lot more. This markup was a local form of a SGML (a generic markup protocol) but the articles there were available only within CERN’s network at the time.
Berners-Lee (now a Sir) thought of using this markup format as a base to format the documents and to allow everything that would be necessary in a science article, but he had also other plans: to allow the articles to be accessible anywhere in the world where a landline and a computer could be available, using a technology that US’s Defense Department had created back in the 1970s in its Advanced Research section (in which was called ARPANET). Further than that, though, he wanted these articles to reference one another and to allow a progressive navigation, in a way we ended up calling a “web” of articles. A protocol was created to index page servers and give them unique addresses, in a way their pages could be accessed only by having their addresses, from anywhere. Also, the base markup language was tweaked and the resulting product, among other changes, featured linking between articles, making these texts now hypertexts — texts that could “expand” indefinitely.
As the idea of a hypertext itself was the obvious difference here, the protocol was called HTTP (Hypertext Transfer Protocol) and the markup was called HTML (Hypertext Markup Language).
As the concept of the now called World Wide Web (named after the only browser available, written by Berners-Lee) extrapolated the realms of the CERN and even the academic domains, it was clear that the technology of the web was going to be used to express much more than what anyone had envisioned at the beginning.
The new non-academic Web users and creators from the early 1990s were demanding more visual possibilities from the HTML, totally crafted for text and little else (see the list of tags of HTML 1.0: not even images or tables were allowed). With that, new additions for the HTML standard were suggested, with low success until Tim Berners-Lee himself founded, in 1994, a committee to take care of Web’s standards, that was called World Wide Web Consortium, or W3C for short. For context, on 1993, with the first commercial Web browser (Mosaic) the Web’s popularity began increasing exponentially. By 1995 the popularity of the Web was already so big that even the giant Microsoft stepped in with its first browser (built on top of Mosaic).
By that time, already with heavy commercial use of the Web, new HTML standards were set up, bringing tables, images, forms more or less as we know them by 1995/96. However, as commercial use of the internet pushed boundaries even further, people struggled to achieve visually impressive pieces having to do hackish layouts, using tables for example (something that people that assemble e-mail markups, tragically, still have to deal with to this very day, in 2021).
The main issue of Web at that point (circa 1995) is that HTML happened to take three different roles to itself: content, presentation and functionality. HTML was created with a very limited scope in mind so there was no problem if it focused on content but took a bit of the other two, because the needs for those were very small. As the Web demanded a lot more possibilities, though, two tools came to release HTML from a burden of dealing clumsily with something it was clearly not made for: CSS for presentation and JavaScript for functionality. Also, as an important side note, security issues became a real concern on the Web, which helped shaping some future functionalities.
One fascinating thing about the web (and the trifecta HTML + CSS + JS) is also one of its biggest downsides: backwards compatibility. The fact we can still read a page from CERN from around the very birth of the Web is amazing, and with minor exceptions (due mostly to security concerns, like Flash pages) all older pages are bound to be readable and even functional, if they are available online. Many pages from the early days of the web simply aren’t hosted anymore, because hosting costs money, but they would most probably work if hosted. There is, though, the other side of the coin: deprecated technologies still have to be supported on the browsers to make this work (again, if there were no security issues attached) and this makes HTML, CSS and JS way more bloated and even a bit less dynamic that they could be if browsers didn’t have to be backwards compatible. HTML, for example, has to carry around tags that are clearly layout tags, or ones that supported deprecated technologies like Java Applets or frame sets. Also, some tags have lacking names and the naming all around feels very inconsistent, but that’s the price to pay to be able to read a website from more than thirty years ago.
All in all, given both the restricted nature of the early Web and the decentralised nature of the later web, we can say that entities like the W3C or companies like Mozilla and even Microsoft, Google and Apple contributed a lot to this state we have now of a fairly good standardisation and a decent environment for Web developers overall. One can say it is not perfect though, but I feel we’re heading towards an even more exciting time ahead in Web development, with growing power for PWAs, WebAssembly and other cool things going on.
Update → we have a sequel to this article just uploaded: The history of deprecated and changed HTML tags. Enjoy!
